diff --git "a/checkpoint-686880/trainer_state.json" "b/checkpoint-686880/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-686880/trainer_state.json"
@@ -0,0 +1,8143 @@
+{
+  "best_metric": 3.868178129196167,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/re-irr-sv-agr/transformer/4/checkpoints/checkpoint-686880",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 686880,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 11.009,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8095,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.1651,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9594,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.7982,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.694,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5853,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.5267,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4291,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.3859,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3256,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.3,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2522,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.1968,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1645,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1134,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.089,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0669,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.0378,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.0085,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 4.9983,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9686,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9489,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9122,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9128,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 4.8816,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 4.8695,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.8503,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8233,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756873037409844e-05,
+      "loss": 4.8105,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748487089899324e-05,
+      "loss": 4.7914,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7841,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7687,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7612,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7491,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7362,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969820416244367e-05,
+      "loss": 4.7235,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.6883,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.6873,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6762,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6602,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6636,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6576,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6361,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6361,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.6288,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6205,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.6164,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5972,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5849,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.5868,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5729,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.5727,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 4.5461,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5598,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.5338,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5539,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5278,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5255,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.518,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.513,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.4991,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.5042,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4797,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4747,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945510995746097e-05,
+      "loss": 4.4883,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944672400995045e-05,
+      "loss": 4.48,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4671,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4529,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4541,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 4.4609,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 4.4558,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.4483,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4451,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4465,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 4.4403,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 4.4182,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 4.42,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 4.4199,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 4.4095,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 4.4097,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 4.4036,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 4.4193,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 4.3935,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 4.3795,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 4.3901,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 4.3872,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 4.3929,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926229867993394e-05,
+      "loss": 4.3719,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925392911122715e-05,
+      "loss": 4.3744,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924554316371663e-05,
+      "loss": 4.3727,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923715721620611e-05,
+      "loss": 4.3668,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922877126869559e-05,
+      "loss": 4.3566,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922038532118507e-05,
+      "loss": 4.3593,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921199937367455e-05,
+      "loss": 4.3604,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920362980496776e-05,
+      "loss": 4.3678,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.3472,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.3402,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.3392,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.3386,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.3461,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.3322,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.3259,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3194,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.3281,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.2985,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.3173,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.3088,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.315,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.3032,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.3084,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.2974,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.2982,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.3028,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.2857,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.3022,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.2988,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901920447495124e-05,
+      "loss": 4.2908,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.2812,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.2801,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.2832,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.2801,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.2792,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.268,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.2754,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.2601,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.2638,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.2515,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.2554,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918605862432463e-05,
+      "loss": 4.2437,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.2666,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.2705,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.2571,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.2336,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.2548,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.2475,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.2457,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885155103995577e-05,
+      "loss": 4.2374,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.2459,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.2392,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.2337,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.2348,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.2369,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.223,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.225,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.237,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.2213,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.2224,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.2328,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.2133,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.207637310028076,
+      "eval_runtime": 302.2209,
+      "eval_samples_per_second": 1262.623,
+      "eval_steps_per_second": 39.458,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.1968,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.2045,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.2192,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.2089,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.2145,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.192,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.2047,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.1838,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867549527864604e-05,
+      "loss": 4.2006,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.1845,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.2007,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.1956,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864196786740769e-05,
+      "loss": 4.1857,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863358191989717e-05,
+      "loss": 4.1773,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862519597238665e-05,
+      "loss": 4.1703,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861681002487613e-05,
+      "loss": 4.1704,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860842407736561e-05,
+      "loss": 4.1794,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.177,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.169,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.1887,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.1814,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.1729,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.164,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.1831,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541369254888915e-05,
+      "loss": 4.1552,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532983307378395e-05,
+      "loss": 4.1677,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.1657,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.1584,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.1492,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.1501,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8491069948629524e-05,
+      "loss": 4.1548,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482684001119e-05,
+      "loss": 4.1533,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847429805360848e-05,
+      "loss": 4.1539,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.1527,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8457526158587444e-05,
+      "loss": 4.1513,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8449140211076924e-05,
+      "loss": 4.1481,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8440754263566404e-05,
+      "loss": 4.1318,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.137,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.1327,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.1284,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.1384,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839885728362126e-05,
+      "loss": 4.1396,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839047133611074e-05,
+      "loss": 4.1273,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838208538860022e-05,
+      "loss": 4.1265,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.1296,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836531349357918e-05,
+      "loss": 4.1312,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835692754606866e-05,
+      "loss": 4.1323,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834854159855814e-05,
+      "loss": 4.1226,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8340188408655086e-05,
+      "loss": 4.1204,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331802461144566e-05,
+      "loss": 4.1243,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323416513634046e-05,
+      "loss": 4.1165,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315030566123526e-05,
+      "loss": 4.1194,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306644618613006e-05,
+      "loss": 4.1097,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298275049906215e-05,
+      "loss": 4.1225,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289889102395695e-05,
+      "loss": 4.0954,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281503154885175e-05,
+      "loss": 4.12,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273117207374655e-05,
+      "loss": 4.1087,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.826474763866787e-05,
+      "loss": 4.106,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825636169115735e-05,
+      "loss": 4.1145,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824797574364683e-05,
+      "loss": 4.1009,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823958979613631e-05,
+      "loss": 4.0953,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823122022742952e-05,
+      "loss": 4.1112,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222834279919e-05,
+      "loss": 4.0843,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821444833240848e-05,
+      "loss": 4.0861,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820606238489796e-05,
+      "loss": 4.1004,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819767643738744e-05,
+      "loss": 4.1008,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818930686868065e-05,
+      "loss": 4.0937,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818092092117013e-05,
+      "loss": 4.0827,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817253497365961e-05,
+      "loss": 4.084,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816414902614909e-05,
+      "loss": 4.0893,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155779457442305e-05,
+      "loss": 4.0975,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147393509931785e-05,
+      "loss": 4.0889,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139007562421265e-05,
+      "loss": 4.0939,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130621614910745e-05,
+      "loss": 4.0931,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122252046203954e-05,
+      "loss": 4.0939,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113866098693434e-05,
+      "loss": 4.075,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105480151182914e-05,
+      "loss": 4.0784,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8097094203672393e-05,
+      "loss": 4.0866,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8088708256161873e-05,
+      "loss": 4.076,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808033868745508e-05,
+      "loss": 4.0787,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807195273994456e-05,
+      "loss": 4.0716,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806356679243404e-05,
+      "loss": 4.0936,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805518084492352e-05,
+      "loss": 4.0728,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804681127621674e-05,
+      "loss": 4.0612,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803842532870622e-05,
+      "loss": 4.0729,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80300393811957e-05,
+      "loss": 4.0741,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802165343368518e-05,
+      "loss": 4.0808,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801326748617466e-05,
+      "loss": 4.0644,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800489791746787e-05,
+      "loss": 4.0654,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799651196995735e-05,
+      "loss": 4.0744,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798812602244683e-05,
+      "loss": 4.0628,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79797400749363e-05,
+      "loss": 4.0584,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7971370506229516e-05,
+      "loss": 4.0626,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7962984558718996e-05,
+      "loss": 4.0703,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954598611208476e-05,
+      "loss": 4.0701,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946212663697956e-05,
+      "loss": 4.0651,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793784309499117e-05,
+      "loss": 4.0488,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792945714748065e-05,
+      "loss": 4.0553,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792107119997013e-05,
+      "loss": 4.0571,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912685252459605e-05,
+      "loss": 4.0613,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790431568375282e-05,
+      "loss": 4.0571,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78959297362423e-05,
+      "loss": 4.0518,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887543788731774e-05,
+      "loss": 4.0439,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879157841221254e-05,
+      "loss": 4.0589,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787078827251447e-05,
+      "loss": 4.026,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786240232500395e-05,
+      "loss": 4.0519,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7854032756297166e-05,
+      "loss": 4.0421,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845646808786646e-05,
+      "loss": 4.0537,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837260861276126e-05,
+      "loss": 4.0394,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828874913765606e-05,
+      "loss": 4.0507,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782048896625508e-05,
+      "loss": 4.0389,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781210301874456e-05,
+      "loss": 4.0395,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803733450037775e-05,
+      "loss": 4.0488,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779534750252725e-05,
+      "loss": 4.036,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778696155501673e-05,
+      "loss": 4.0485,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777857560750621e-05,
+      "loss": 4.0522,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7770206038799423e-05,
+      "loss": 4.0434,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7761820091288903e-05,
+      "loss": 4.0349,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7753434143778383e-05,
+      "loss": 4.0379,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774504819626786e-05,
+      "loss": 4.0406,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773666224875734e-05,
+      "loss": 4.0287,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772827630124682e-05,
+      "loss": 4.0446,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77198903537363e-05,
+      "loss": 4.0282,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771150440622578e-05,
+      "loss": 4.0365,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770313483751899e-05,
+      "loss": 4.0311,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769474889000847e-05,
+      "loss": 4.027,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768636294249795e-05,
+      "loss": 4.0172,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767797699498743e-05,
+      "loss": 4.0242,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766960742628064e-05,
+      "loss": 4.016,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766122147877012e-05,
+      "loss": 4.037,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76528355312596e-05,
+      "loss": 4.0458,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764444958374909e-05,
+      "loss": 4.0317,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76360800150423e-05,
+      "loss": 4.0103,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762769406753178e-05,
+      "loss": 4.0325,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761930812002126e-05,
+      "loss": 4.0236,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761092217251074e-05,
+      "loss": 4.0272,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602552603803946e-05,
+      "loss": 4.0224,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594166656293426e-05,
+      "loss": 4.0245,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7585780708782906e-05,
+      "loss": 4.0196,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7577394761272386e-05,
+      "loss": 4.0208,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569025192565595e-05,
+      "loss": 4.0191,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560639245055075e-05,
+      "loss": 4.0241,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552253297544555e-05,
+      "loss": 4.0166,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754388372883777e-05,
+      "loss": 4.0099,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753549778132725e-05,
+      "loss": 4.0266,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752711183381673e-05,
+      "loss": 4.0186,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751872588630621e-05,
+      "loss": 4.0112,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751033993879569e-05,
+      "loss": 4.0261,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750195399128517e-05,
+      "loss": 4.0078,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.040642738342285,
+      "eval_runtime": 301.9306,
+      "eval_samples_per_second": 1263.837,
+      "eval_steps_per_second": 39.496,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749356804377465e-05,
+      "loss": 3.9944,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748518209626413e-05,
+      "loss": 4.0006,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747679614875361e-05,
+      "loss": 4.0196,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468410201243084e-05,
+      "loss": 4.012,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460024253732563e-05,
+      "loss": 4.0153,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451638306222043e-05,
+      "loss": 3.9956,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744325235871152e-05,
+      "loss": 4.0103,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434866411201e-05,
+      "loss": 3.9869,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742648046369049e-05,
+      "loss": 4.0128,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741809451617997e-05,
+      "loss": 3.9934,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740970856866945e-05,
+      "loss": 4.01,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740133899996266e-05,
+      "loss": 4.004,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739296943125587e-05,
+      "loss": 3.9969,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738458348374535e-05,
+      "loss": 3.9886,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737619753623483e-05,
+      "loss": 3.9864,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736781158872431e-05,
+      "loss": 3.9868,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735944202001752e-05,
+      "loss": 3.9928,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7351056072507e-05,
+      "loss": 3.9946,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734267012499648e-05,
+      "loss": 3.9913,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733428417748596e-05,
+      "loss": 4.0051,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325898229975444e-05,
+      "loss": 4.0017,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317512282464924e-05,
+      "loss": 3.9974,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7309126334954404e-05,
+      "loss": 3.9895,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300740387443884e-05,
+      "loss": 4.001,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729237081873709e-05,
+      "loss": 3.9816,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728398487122657e-05,
+      "loss": 3.993,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727559892371605e-05,
+      "loss": 3.9909,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726721297620553e-05,
+      "loss": 3.9876,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725884340749874e-05,
+      "loss": 3.9755,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725045745998822e-05,
+      "loss": 3.9784,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72420715124777e-05,
+      "loss": 3.9816,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723370194377091e-05,
+      "loss": 3.9881,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72253159962604e-05,
+      "loss": 3.9895,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721693004874988e-05,
+      "loss": 3.9817,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720854410123936e-05,
+      "loss": 3.9821,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720015815372884e-05,
+      "loss": 3.9811,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719177220621832e-05,
+      "loss": 3.9668,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71833862587078e-05,
+      "loss": 3.977,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7175016690001006e-05,
+      "loss": 3.9696,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7166630742490486e-05,
+      "loss": 3.9687,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7158244794979966e-05,
+      "loss": 3.9724,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7149858847469446e-05,
+      "loss": 3.9813,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141489278762655e-05,
+      "loss": 3.9687,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133103331252135e-05,
+      "loss": 3.9645,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124717383741615e-05,
+      "loss": 3.9778,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116331436231095e-05,
+      "loss": 3.97,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710796186752431e-05,
+      "loss": 3.9781,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709957592001379e-05,
+      "loss": 3.9639,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709118997250327e-05,
+      "loss": 3.9609,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082804024992744e-05,
+      "loss": 3.9711,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707443445628596e-05,
+      "loss": 3.9637,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706604850877544e-05,
+      "loss": 3.9671,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705767894006865e-05,
+      "loss": 3.9577,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704929299255813e-05,
+      "loss": 3.9637,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704090704504761e-05,
+      "loss": 3.9505,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703252109753709e-05,
+      "loss": 3.9655,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702413515002657e-05,
+      "loss": 3.96,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701574920251605e-05,
+      "loss": 3.9593,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700736325500553e-05,
+      "loss": 3.9683,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699897730749501e-05,
+      "loss": 3.9509,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699060773878822e-05,
+      "loss": 3.9473,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69822217912777e-05,
+      "loss": 3.9686,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697383584376718e-05,
+      "loss": 3.9398,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696544989625666e-05,
+      "loss": 3.939,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695708032754987e-05,
+      "loss": 3.9541,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694869438003935e-05,
+      "loss": 3.9597,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694032481133256e-05,
+      "loss": 3.9513,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693193886382204e-05,
+      "loss": 3.9421,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692355291631152e-05,
+      "loss": 3.9432,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6915166968801e-05,
+      "loss": 3.944,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690679740009422e-05,
+      "loss": 3.9584,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689841145258369e-05,
+      "loss": 3.9506,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689002550507317e-05,
+      "loss": 3.9533,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688163955756265e-05,
+      "loss": 3.9523,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687325361005213e-05,
+      "loss": 3.957,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686486766254161e-05,
+      "loss": 3.9378,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685648171503109e-05,
+      "loss": 3.9414,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684809576752057e-05,
+      "loss": 3.9504,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683974257761751e-05,
+      "loss": 3.9362,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831356630106996e-05,
+      "loss": 3.9446,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822970682596476e-05,
+      "loss": 3.9367,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814584735085956e-05,
+      "loss": 3.9553,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806198787575436e-05,
+      "loss": 3.9401,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797812840064916e-05,
+      "loss": 3.9304,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789426892554396e-05,
+      "loss": 3.9352,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6781040945043876e-05,
+      "loss": 3.9419,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772671376337085e-05,
+      "loss": 3.9525,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764285428826565e-05,
+      "loss": 3.9335,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755899481316045e-05,
+      "loss": 3.9372,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747513533805525e-05,
+      "loss": 3.9351,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739127586295005e-05,
+      "loss": 3.9426,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730758017588214e-05,
+      "loss": 3.9304,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722372070077694e-05,
+      "loss": 3.9296,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713986122567174e-05,
+      "loss": 3.9414,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670560017505666e-05,
+      "loss": 3.9424,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669723060634987e-05,
+      "loss": 3.9398,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668884465883935e-05,
+      "loss": 3.9225,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668045871132883e-05,
+      "loss": 3.9294,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667207276381831e-05,
+      "loss": 3.9302,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666368681630779e-05,
+      "loss": 3.9343,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665530086879727e-05,
+      "loss": 3.9345,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664693130009048e-05,
+      "loss": 3.9256,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663854535257996e-05,
+      "loss": 3.9216,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663015940506944e-05,
+      "loss": 3.9341,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662177345755892e-05,
+      "loss": 3.9076,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66133875100484e-05,
+      "loss": 3.9262,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6605017941341614e-05,
+      "loss": 3.9179,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596631993831094e-05,
+      "loss": 3.9324,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588246046320574e-05,
+      "loss": 3.9218,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579860098810054e-05,
+      "loss": 3.9266,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657147415129953e-05,
+      "loss": 3.9184,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656308820378901e-05,
+      "loss": 3.9211,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655471863508222e-05,
+      "loss": 3.9287,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65463326875717e-05,
+      "loss": 3.9201,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537946740061176e-05,
+      "loss": 3.9297,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529560792550656e-05,
+      "loss": 3.9314,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65212076026476e-05,
+      "loss": 3.9286,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651282165513708e-05,
+      "loss": 3.9192,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650443570762657e-05,
+      "loss": 3.9149,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649604976011605e-05,
+      "loss": 3.9276,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648766381260553e-05,
+      "loss": 3.9112,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479277865095e-05,
+      "loss": 3.9256,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647089191758448e-05,
+      "loss": 3.9139,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646250597007396e-05,
+      "loss": 3.921,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645413640136718e-05,
+      "loss": 3.9166,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644575045385665e-05,
+      "loss": 3.9148,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643736450634613e-05,
+      "loss": 3.9028,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642897855883561e-05,
+      "loss": 3.9131,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642059261132509e-05,
+      "loss": 3.9072,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6412223042618306e-05,
+      "loss": 3.9176,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6403837095107786e-05,
+      "loss": 3.9305,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395451147597266e-05,
+      "loss": 3.9247,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6387065200086746e-05,
+      "loss": 3.8943,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378679252576226e-05,
+      "loss": 3.9243,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370309683869435e-05,
+      "loss": 3.9136,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361923736358915e-05,
+      "loss": 3.9142,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353537788848395e-05,
+      "loss": 3.9034,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6345151841337874e-05,
+      "loss": 3.9161,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6336765893827354e-05,
+      "loss": 3.9085,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6328379946316834e-05,
+      "loss": 3.9107,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319993998806314e-05,
+      "loss": 3.912,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311624430099523e-05,
+      "loss": 3.9148,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303238482589003e-05,
+      "loss": 3.9099,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629486891388222e-05,
+      "loss": 3.9053,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62864829663717e-05,
+      "loss": 3.9144,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627809701886118e-05,
+      "loss": 3.9132,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626971107135066e-05,
+      "loss": 3.9037,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626132512384014e-05,
+      "loss": 3.9176,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625293917632962e-05,
+      "loss": 3.9019,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9672722816467285,
+      "eval_runtime": 325.6205,
+      "eval_samples_per_second": 1171.889,
+      "eval_steps_per_second": 36.622,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62445532288191e-05,
+      "loss": 3.894,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623616728130858e-05,
+      "loss": 3.8961,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622778133379806e-05,
+      "loss": 3.9129,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621939538628753e-05,
+      "loss": 3.9046,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621100943877701e-05,
+      "loss": 3.9166,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620263987007023e-05,
+      "loss": 3.8888,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619425392255971e-05,
+      "loss": 3.9068,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618586797504919e-05,
+      "loss": 3.8817,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617748202753867e-05,
+      "loss": 3.9123,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6169112458831884e-05,
+      "loss": 3.8929,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160726511321364e-05,
+      "loss": 3.9041,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615234056381084e-05,
+      "loss": 3.9026,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614397099510405e-05,
+      "loss": 3.8895,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613558504759353e-05,
+      "loss": 3.8939,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127199100083006e-05,
+      "loss": 3.8864,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118813152572486e-05,
+      "loss": 3.8845,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110427205061966e-05,
+      "loss": 3.8955,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6102041257551446e-05,
+      "loss": 3.8922,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093655310040926e-05,
+      "loss": 3.8918,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085269362530406e-05,
+      "loss": 3.9051,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607689979382362e-05,
+      "loss": 3.9054,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60685138463131e-05,
+      "loss": 3.8967,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606012789880258e-05,
+      "loss": 3.8917,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605174195129206e-05,
+      "loss": 3.9009,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604337238258527e-05,
+      "loss": 3.8846,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603498643507475e-05,
+      "loss": 3.8982,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602660048756423e-05,
+      "loss": 3.8931,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601821454005371e-05,
+      "loss": 3.8865,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600984497134692e-05,
+      "loss": 3.8795,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.60014590238364e-05,
+      "loss": 3.8827,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599307307632588e-05,
+      "loss": 3.8853,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598468712881536e-05,
+      "loss": 3.893,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976317560108575e-05,
+      "loss": 3.8944,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967931612598055e-05,
+      "loss": 3.887,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959545665087535e-05,
+      "loss": 3.8871,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5951159717577015e-05,
+      "loss": 3.8875,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942773770066495e-05,
+      "loss": 3.8725,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934404201359704e-05,
+      "loss": 3.8812,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5926018253849184e-05,
+      "loss": 3.8721,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917632306338664e-05,
+      "loss": 3.8776,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909246358828144e-05,
+      "loss": 3.8785,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590087679012135e-05,
+      "loss": 3.8837,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589249084261083e-05,
+      "loss": 3.8836,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588410489510031e-05,
+      "loss": 3.8703,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58757189475898e-05,
+      "loss": 3.8834,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586733300007928e-05,
+      "loss": 3.884,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585896343137249e-05,
+      "loss": 3.884,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585057748386197e-05,
+      "loss": 3.8757,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584219153635145e-05,
+      "loss": 3.8671,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583380558884093e-05,
+      "loss": 3.8814,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582543602013414e-05,
+      "loss": 3.8715,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581705007262362e-05,
+      "loss": 3.877,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58086641251131e-05,
+      "loss": 3.8691,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580027817760258e-05,
+      "loss": 3.8701,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579189223009206e-05,
+      "loss": 3.8651,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783522661385267e-05,
+      "loss": 3.8755,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577513671387475e-05,
+      "loss": 3.8724,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576675076636423e-05,
+      "loss": 3.873,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575836481885371e-05,
+      "loss": 3.8763,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574999525014692e-05,
+      "loss": 3.8612,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57416093026364e-05,
+      "loss": 3.8626,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573322335512588e-05,
+      "loss": 3.8834,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572483740761536e-05,
+      "loss": 3.8524,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571646783890857e-05,
+      "loss": 3.8507,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570808189139805e-05,
+      "loss": 3.8635,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569969594388753e-05,
+      "loss": 3.8745,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569130999637701e-05,
+      "loss": 3.8643,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568294042767022e-05,
+      "loss": 3.8594,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567455448015971e-05,
+      "loss": 3.8538,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566616853264919e-05,
+      "loss": 3.8587,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565778258513867e-05,
+      "loss": 3.8734,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649413016431876e-05,
+      "loss": 3.8655,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5641027068921356e-05,
+      "loss": 3.8638,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632641121410836e-05,
+      "loss": 3.8721,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624255173900316e-05,
+      "loss": 3.8721,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561586922638979e-05,
+      "loss": 3.8536,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607499657683005e-05,
+      "loss": 3.86,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599113710172485e-05,
+      "loss": 3.8673,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590727762661965e-05,
+      "loss": 3.8492,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582341815151445e-05,
+      "loss": 3.8622,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5573955867640925e-05,
+      "loss": 3.8542,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556558629893414e-05,
+      "loss": 3.8734,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557200351423614e-05,
+      "loss": 3.8577,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548814403913094e-05,
+      "loss": 3.8496,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540428456402574e-05,
+      "loss": 3.848,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553205888769579e-05,
+      "loss": 3.8589,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552367294018526e-05,
+      "loss": 3.8698,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551528699267474e-05,
+      "loss": 3.855,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550690104516422e-05,
+      "loss": 3.852,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549853147645744e-05,
+      "loss": 3.8514,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549014552894691e-05,
+      "loss": 3.8635,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548175958143639e-05,
+      "loss": 3.8487,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547337363392588e-05,
+      "loss": 3.8507,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546500406521909e-05,
+      "loss": 3.8608,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545661811770857e-05,
+      "loss": 3.8642,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544823217019805e-05,
+      "loss": 3.8561,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543984622268753e-05,
+      "loss": 3.8402,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431476653980736e-05,
+      "loss": 3.8513,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5423090706470216e-05,
+      "loss": 3.8487,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414704758959696e-05,
+      "loss": 3.8552,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406318811449176e-05,
+      "loss": 3.8549,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397949242742385e-05,
+      "loss": 3.8464,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389563295231865e-05,
+      "loss": 3.8449,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5381177347721345e-05,
+      "loss": 3.8519,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537279140021083e-05,
+      "loss": 3.832,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536442183150404e-05,
+      "loss": 3.8483,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535603588399352e-05,
+      "loss": 3.8372,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347649936483e-05,
+      "loss": 3.8548,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533926398897248e-05,
+      "loss": 3.8441,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533089442026569e-05,
+      "loss": 3.8458,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532250847275517e-05,
+      "loss": 3.8409,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531412252524465e-05,
+      "loss": 3.8474,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530573657773413e-05,
+      "loss": 3.8524,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529736700902734e-05,
+      "loss": 3.8391,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528898106151682e-05,
+      "loss": 3.8509,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52805951140063e-05,
+      "loss": 3.8542,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5272225545299515e-05,
+      "loss": 3.8489,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263839597788995e-05,
+      "loss": 3.8486,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255453650278475e-05,
+      "loss": 3.8358,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5247067702767955e-05,
+      "loss": 3.8526,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238681755257435e-05,
+      "loss": 3.841,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230295807746915e-05,
+      "loss": 3.8461,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5221909860236395e-05,
+      "loss": 3.8355,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5213523912725875e-05,
+      "loss": 3.8454,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5205154344019084e-05,
+      "loss": 3.8409,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196768396508564e-05,
+      "loss": 3.8375,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5188382448998044e-05,
+      "loss": 3.8301,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5179996501487524e-05,
+      "loss": 3.8345,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517162693278074e-05,
+      "loss": 3.8357,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516324098527022e-05,
+      "loss": 3.8417,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51548550377597e-05,
+      "loss": 3.8522,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514646909024918e-05,
+      "loss": 3.8514,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513809952154239e-05,
+      "loss": 3.8224,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512971357403187e-05,
+      "loss": 3.8519,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512132762652135e-05,
+      "loss": 3.8389,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511294167901083e-05,
+      "loss": 3.8404,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510457211030404e-05,
+      "loss": 3.8313,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509618616279352e-05,
+      "loss": 3.842,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087800215283e-05,
+      "loss": 3.8342,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507941426777248e-05,
+      "loss": 3.8441,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507104469906569e-05,
+      "loss": 3.8361,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506265875155517e-05,
+      "loss": 3.8396,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505427280404465e-05,
+      "loss": 3.8406,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504588685653413e-05,
+      "loss": 3.8312,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503751728782734e-05,
+      "loss": 3.84,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502913134031682e-05,
+      "loss": 3.8458,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50207453928063e-05,
+      "loss": 3.8311,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501235944529578e-05,
+      "loss": 3.8401,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500397349778526e-05,
+      "loss": 3.8342,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9272706508636475,
+      "eval_runtime": 302.1874,
+      "eval_samples_per_second": 1262.763,
+      "eval_steps_per_second": 39.462,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499558755027474e-05,
+      "loss": 3.8199,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987201602764215e-05,
+      "loss": 3.8236,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978815655253695e-05,
+      "loss": 3.8368,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970429707743175e-05,
+      "loss": 3.8359,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496204376023266e-05,
+      "loss": 3.848,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495365781272214e-05,
+      "loss": 3.8183,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494527186521162e-05,
+      "loss": 3.8324,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49368859177011e-05,
+      "loss": 3.8182,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492851634899431e-05,
+      "loss": 3.8384,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492013040148379e-05,
+      "loss": 3.8232,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491174445397327e-05,
+      "loss": 3.8298,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490335850646275e-05,
+      "loss": 3.8359,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489500531655969e-05,
+      "loss": 3.8188,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488661936904917e-05,
+      "loss": 3.8292,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487823342153865e-05,
+      "loss": 3.8188,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486984747402813e-05,
+      "loss": 3.8148,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486146152651761e-05,
+      "loss": 3.8223,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4853075579007095e-05,
+      "loss": 3.8259,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844689631496575e-05,
+      "loss": 3.8231,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4836303683986055e-05,
+      "loss": 3.8354,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827934115279264e-05,
+      "loss": 3.8381,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4819548167768744e-05,
+      "loss": 3.829,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4811162220258224e-05,
+      "loss": 3.8263,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4802776272747704e-05,
+      "loss": 3.8342,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479440670404091e-05,
+      "loss": 3.8144,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478602075653039e-05,
+      "loss": 3.8315,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477763480901987e-05,
+      "loss": 3.8272,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476924886150935e-05,
+      "loss": 3.8178,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476087929280256e-05,
+      "loss": 3.8149,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475249334529205e-05,
+      "loss": 3.8122,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474410739778153e-05,
+      "loss": 3.8191,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473572145027101e-05,
+      "loss": 3.8248,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472735188156422e-05,
+      "loss": 3.8281,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47189659340537e-05,
+      "loss": 3.8222,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471057998654318e-05,
+      "loss": 3.8228,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470219403903266e-05,
+      "loss": 3.8213,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469382447032587e-05,
+      "loss": 3.8045,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468543852281535e-05,
+      "loss": 3.8186,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467705257530483e-05,
+      "loss": 3.8069,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466866662779431e-05,
+      "loss": 3.8117,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660297059087516e-05,
+      "loss": 3.8107,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651911111577e-05,
+      "loss": 3.8256,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464352516406648e-05,
+      "loss": 3.8174,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463513921655596e-05,
+      "loss": 3.8043,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462676964784917e-05,
+      "loss": 3.8215,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461838370033865e-05,
+      "loss": 3.8165,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460999775282813e-05,
+      "loss": 3.8197,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460161180531761e-05,
+      "loss": 3.8123,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459324223661082e-05,
+      "loss": 3.8021,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45848562891003e-05,
+      "loss": 3.8196,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457647034158978e-05,
+      "loss": 3.8066,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456808439407926e-05,
+      "loss": 3.8145,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455971482537247e-05,
+      "loss": 3.8056,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551328877861956e-05,
+      "loss": 3.8048,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542942930351436e-05,
+      "loss": 3.8041,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534556982840916e-05,
+      "loss": 3.8106,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4526187414134125e-05,
+      "loss": 3.8077,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517801466623605e-05,
+      "loss": 3.8092,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509415519113085e-05,
+      "loss": 3.8165,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4501029571602565e-05,
+      "loss": 3.7997,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492660002895774e-05,
+      "loss": 3.7971,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4484274055385254e-05,
+      "loss": 3.8236,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475888107874734e-05,
+      "loss": 3.7875,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467502160364214e-05,
+      "loss": 3.7889,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445913259165742e-05,
+      "loss": 3.8015,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445074664414691e-05,
+      "loss": 3.816,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444236069663639e-05,
+      "loss": 3.7973,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443397474912587e-05,
+      "loss": 3.8023,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442560518041908e-05,
+      "loss": 3.7931,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441721923290856e-05,
+      "loss": 3.7987,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440883328539804e-05,
+      "loss": 3.808,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440044733788752e-05,
+      "loss": 3.806,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439207776918073e-05,
+      "loss": 3.8031,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438369182167021e-05,
+      "loss": 3.8081,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437530587415969e-05,
+      "loss": 3.8101,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436691992664917e-05,
+      "loss": 3.7962,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435855035794238e-05,
+      "loss": 3.7975,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4350164410431864e-05,
+      "loss": 3.8091,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341778462921344e-05,
+      "loss": 3.7866,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433340889421455e-05,
+      "loss": 3.8009,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432502294670403e-05,
+      "loss": 3.7934,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431663699919351e-05,
+      "loss": 3.8139,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430825105168299e-05,
+      "loss": 3.7983,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.42998814829762e-05,
+      "loss": 3.7925,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429149553546568e-05,
+      "loss": 3.7823,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428310958795516e-05,
+      "loss": 3.8034,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427472364044464e-05,
+      "loss": 3.8059,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426633769293412e-05,
+      "loss": 3.7998,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425796812422733e-05,
+      "loss": 3.792,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424958217671682e-05,
+      "loss": 3.7952,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42411962292063e-05,
+      "loss": 3.8023,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232826660499506e-05,
+      "loss": 3.7905,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224440712988986e-05,
+      "loss": 3.79,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4216054765478466e-05,
+      "loss": 3.8023,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207668817967946e-05,
+      "loss": 3.8027,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199299249261155e-05,
+      "loss": 3.7995,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190913301750635e-05,
+      "loss": 3.7843,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182527354240115e-05,
+      "loss": 3.7936,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174141406729595e-05,
+      "loss": 3.7878,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165771838022804e-05,
+      "loss": 3.7957,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157385890512284e-05,
+      "loss": 3.7962,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414899994300177e-05,
+      "loss": 3.7936,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140613995491244e-05,
+      "loss": 3.7868,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413224442678446e-05,
+      "loss": 3.7887,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412385847927394e-05,
+      "loss": 3.7823,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411547253176342e-05,
+      "loss": 3.7869,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410708658425289e-05,
+      "loss": 3.7802,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409871701554611e-05,
+      "loss": 3.797,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409033106803559e-05,
+      "loss": 3.789,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408194512052507e-05,
+      "loss": 3.7876,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407355917301454e-05,
+      "loss": 3.7825,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406518960430776e-05,
+      "loss": 3.7937,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405680365679724e-05,
+      "loss": 3.7929,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404841770928672e-05,
+      "loss": 3.7819,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40400317617762e-05,
+      "loss": 3.7948,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031662193069414e-05,
+      "loss": 3.797,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402329262436262e-05,
+      "loss": 3.7918,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40149066768521e-05,
+      "loss": 3.7934,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400652072934158e-05,
+      "loss": 3.7798,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399813478183106e-05,
+      "loss": 3.7919,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398974883432054e-05,
+      "loss": 3.7868,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981362886810016e-05,
+      "loss": 3.7929,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972976939299496e-05,
+      "loss": 3.7729,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964590991788976e-05,
+      "loss": 3.7918,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395622142308219e-05,
+      "loss": 3.7862,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394783547557167e-05,
+      "loss": 3.7809,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393944952806115e-05,
+      "loss": 3.7758,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393106358055063e-05,
+      "loss": 3.7783,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392269401184384e-05,
+      "loss": 3.7787,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391430806433332e-05,
+      "loss": 3.7915,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39059221168228e-05,
+      "loss": 3.7927,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389753616931228e-05,
+      "loss": 3.7976,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388916660060549e-05,
+      "loss": 3.7693,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388078065309497e-05,
+      "loss": 3.7923,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387239470558445e-05,
+      "loss": 3.787,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386400875807393e-05,
+      "loss": 3.7809,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855639189367145e-05,
+      "loss": 3.7771,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3847253241856625e-05,
+      "loss": 3.7906,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838867294346105e-05,
+      "loss": 3.7811,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830481346835585e-05,
+      "loss": 3.7872,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3822111778128794e-05,
+      "loss": 3.7796,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813725830618274e-05,
+      "loss": 3.7854,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805339883107754e-05,
+      "loss": 3.7869,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796953935597234e-05,
+      "loss": 3.7802,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378858436689044e-05,
+      "loss": 3.7852,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378019841937992e-05,
+      "loss": 3.7919,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37718124718694e-05,
+      "loss": 3.7802,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376342652435888e-05,
+      "loss": 3.7825,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375504057684836e-05,
+      "loss": 3.7842,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.90229868888855,
+      "eval_runtime": 304.1721,
+      "eval_samples_per_second": 1254.523,
+      "eval_steps_per_second": 39.205,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374665462933785e-05,
+      "loss": 3.7724,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373826868182733e-05,
+      "loss": 3.7685,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372989911312054e-05,
+      "loss": 3.7844,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372151316561002e-05,
+      "loss": 3.7843,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37131272180995e-05,
+      "loss": 3.7905,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370474127058898e-05,
+      "loss": 3.7715,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369637170188219e-05,
+      "loss": 3.7777,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368798575437167e-05,
+      "loss": 3.7655,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367959980686115e-05,
+      "loss": 3.783,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367121385935063e-05,
+      "loss": 3.7749,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366284429064384e-05,
+      "loss": 3.7752,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365445834313332e-05,
+      "loss": 3.7806,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364608877442653e-05,
+      "loss": 3.7712,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363770282691601e-05,
+      "loss": 3.7737,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362931687940549e-05,
+      "loss": 3.7672,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362093093189497e-05,
+      "loss": 3.7623,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361254498438445e-05,
+      "loss": 3.7664,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360415903687393e-05,
+      "loss": 3.7733,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359577308936341e-05,
+      "loss": 3.7711,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358738714185289e-05,
+      "loss": 3.7867,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35790175731461e-05,
+      "loss": 3.7854,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357063162563558e-05,
+      "loss": 3.7783,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356224567812506e-05,
+      "loss": 3.7753,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355385973061454e-05,
+      "loss": 3.7806,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354549016190776e-05,
+      "loss": 3.7649,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353710421439724e-05,
+      "loss": 3.7782,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352871826688672e-05,
+      "loss": 3.775,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35203323193762e-05,
+      "loss": 3.7656,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3511962750669406e-05,
+      "loss": 3.7639,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503576803158886e-05,
+      "loss": 3.7647,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3495190855648366e-05,
+      "loss": 3.7695,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3486804908137846e-05,
+      "loss": 3.7734,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478435339431055e-05,
+      "loss": 3.7813,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3470049391920535e-05,
+      "loss": 3.7707,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461663444410015e-05,
+      "loss": 3.7712,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3453277496899495e-05,
+      "loss": 3.7722,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344490792819271e-05,
+      "loss": 3.7599,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343652198068219e-05,
+      "loss": 3.7612,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342813603317167e-05,
+      "loss": 3.7556,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341975008566115e-05,
+      "loss": 3.76,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341138051695436e-05,
+      "loss": 3.7602,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340299456944384e-05,
+      "loss": 3.7766,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339460862193332e-05,
+      "loss": 3.7682,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338622267442279e-05,
+      "loss": 3.7547,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337785310571601e-05,
+      "loss": 3.776,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336946715820549e-05,
+      "loss": 3.7624,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336108121069497e-05,
+      "loss": 3.7745,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335271164198818e-05,
+      "loss": 3.7613,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3344325694477664e-05,
+      "loss": 3.7525,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3335939746967144e-05,
+      "loss": 3.7684,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327553799456624e-05,
+      "loss": 3.7611,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331918423074983e-05,
+      "loss": 3.7688,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331079828323931e-05,
+      "loss": 3.7545,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330241233572879e-05,
+      "loss": 3.7581,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3294026388218266e-05,
+      "loss": 3.7528,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328565681951148e-05,
+      "loss": 3.7628,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327727087200096e-05,
+      "loss": 3.7632,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326888492449044e-05,
+      "loss": 3.7581,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3260498976979915e-05,
+      "loss": 3.7704,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325212940827313e-05,
+      "loss": 3.754,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324374346076261e-05,
+      "loss": 3.7492,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32353575132521e-05,
+      "loss": 3.776,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322697156574157e-05,
+      "loss": 3.7425,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3218618375838516e-05,
+      "loss": 3.7379,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3210232428327996e-05,
+      "loss": 3.7555,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201846480817476e-05,
+      "loss": 3.7691,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193460533306956e-05,
+      "loss": 3.7461,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3185074585796436e-05,
+      "loss": 3.7586,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176688638285916e-05,
+      "loss": 3.7435,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316830269077539e-05,
+      "loss": 3.7502,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315991674326487e-05,
+      "loss": 3.7611,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151547174558085e-05,
+      "loss": 3.7591,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3143161227047565e-05,
+      "loss": 3.7576,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3134775279537045e-05,
+      "loss": 3.7567,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3126389332026525e-05,
+      "loss": 3.7661,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311801976331974e-05,
+      "loss": 3.7475,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109633815809214e-05,
+      "loss": 3.7554,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3101247868298694e-05,
+      "loss": 3.7576,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3092861920788174e-05,
+      "loss": 3.741,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308449235208139e-05,
+      "loss": 3.7536,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307610640457086e-05,
+      "loss": 3.7506,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306772045706034e-05,
+      "loss": 3.7623,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305933450954982e-05,
+      "loss": 3.7566,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305096494084304e-05,
+      "loss": 3.7453,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304257899333252e-05,
+      "loss": 3.737,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3034193045822e-05,
+      "loss": 3.7549,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302580709831148e-05,
+      "loss": 3.76,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301743752960469e-05,
+      "loss": 3.7472,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300905158209417e-05,
+      "loss": 3.7511,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300066563458365e-05,
+      "loss": 3.7483,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299227968707313e-05,
+      "loss": 3.7535,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983910118366337e-05,
+      "loss": 3.7438,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975524170855817e-05,
+      "loss": 3.7429,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2967138223345296e-05,
+      "loss": 3.7597,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958752275834776e-05,
+      "loss": 3.7579,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295038270712799e-05,
+      "loss": 3.7532,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294199675961747e-05,
+      "loss": 3.7383,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293361081210695e-05,
+      "loss": 3.7499,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292522486459643e-05,
+      "loss": 3.7393,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291685529588964e-05,
+      "loss": 3.7538,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290846934837912e-05,
+      "loss": 3.7501,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290009977967234e-05,
+      "loss": 3.748,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289171383216181e-05,
+      "loss": 3.7417,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288332788465129e-05,
+      "loss": 3.7396,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287494193714077e-05,
+      "loss": 3.7443,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286655598963025e-05,
+      "loss": 3.7383,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285817004211973e-05,
+      "loss": 3.7375,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284978409460921e-05,
+      "loss": 3.7498,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28413981470987e-05,
+      "loss": 3.7471,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2833028578391906e-05,
+      "loss": 3.7425,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2824642630881386e-05,
+      "loss": 3.7358,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2816256683370866e-05,
+      "loss": 3.7481,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807887114664075e-05,
+      "loss": 3.7505,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799501167153555e-05,
+      "loss": 3.7386,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2791115219643035e-05,
+      "loss": 3.7497,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782729272132515e-05,
+      "loss": 3.7513,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2774343324621995e-05,
+      "loss": 3.7518,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2765973755915204e-05,
+      "loss": 3.7443,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757587808404684e-05,
+      "loss": 3.7402,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2749201860894164e-05,
+      "loss": 3.744,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274081591338365e-05,
+      "loss": 3.743,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273242996587313e-05,
+      "loss": 3.7479,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272404401836261e-05,
+      "loss": 3.7325,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271565807085209e-05,
+      "loss": 3.7444,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27072885021453e-05,
+      "loss": 3.743,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269890255463478e-05,
+      "loss": 3.7368,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269051660712426e-05,
+      "loss": 3.7296,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268213065961374e-05,
+      "loss": 3.7388,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267376109090695e-05,
+      "loss": 3.7316,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266537514339643e-05,
+      "loss": 3.7453,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265698919588591e-05,
+      "loss": 3.7474,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264860324837539e-05,
+      "loss": 3.7551,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2640233679668604e-05,
+      "loss": 3.7269,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631847732158084e-05,
+      "loss": 3.7484,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262347816345129e-05,
+      "loss": 3.7441,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261509221594077e-05,
+      "loss": 3.738,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260670626843025e-05,
+      "loss": 3.7334,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259832032091973e-05,
+      "loss": 3.7487,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258993437340921e-05,
+      "loss": 3.7332,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258154842589869e-05,
+      "loss": 3.7479,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257316247838817e-05,
+      "loss": 3.734,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564776530877646e-05,
+      "loss": 3.745,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255640696217086e-05,
+      "loss": 3.7417,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254802101466034e-05,
+      "loss": 3.7374,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253965144595356e-05,
+      "loss": 3.7466,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253126549844304e-05,
+      "loss": 3.7423,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252287955093252e-05,
+      "loss": 3.745,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514493603422e-05,
+      "loss": 3.7306,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250610765591147e-05,
+      "loss": 3.7453,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8876891136169434,
+      "eval_runtime": 303.0714,
+      "eval_samples_per_second": 1259.079,
+      "eval_steps_per_second": 39.347,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249772170840095e-05,
+      "loss": 3.7321,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248933576089043e-05,
+      "loss": 3.724,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248094981337991e-05,
+      "loss": 3.7398,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247256386586939e-05,
+      "loss": 3.7432,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246417791835887e-05,
+      "loss": 3.7474,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245579197084835e-05,
+      "loss": 3.7286,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244742240214156e-05,
+      "loss": 3.7341,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243903645463104e-05,
+      "loss": 3.7226,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243065050712052e-05,
+      "loss": 3.742,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2422264559610006e-05,
+      "loss": 3.7316,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413878612099486e-05,
+      "loss": 3.7376,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405492664588966e-05,
+      "loss": 3.7371,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2397123095882175e-05,
+      "loss": 3.7273,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388737148371655e-05,
+      "loss": 3.7364,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2380351200861135e-05,
+      "loss": 3.7217,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371965253350615e-05,
+      "loss": 3.7202,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363579305840095e-05,
+      "loss": 3.7255,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2355193358329575e-05,
+      "loss": 3.7376,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346807410819055e-05,
+      "loss": 3.7278,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233842146330853e-05,
+      "loss": 3.7464,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2330051894601744e-05,
+      "loss": 3.7416,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321665947091224e-05,
+      "loss": 3.7395,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2313279999580704e-05,
+      "loss": 3.7345,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304894052070184e-05,
+      "loss": 3.7353,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22965244833634e-05,
+      "loss": 3.7226,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228813853585288e-05,
+      "loss": 3.7413,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227975258834236e-05,
+      "loss": 3.7305,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227136664083183e-05,
+      "loss": 3.7247,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226299707212505e-05,
+      "loss": 3.7256,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225462750341826e-05,
+      "loss": 3.7253,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224624155590774e-05,
+      "loss": 3.7279,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223785560839722e-05,
+      "loss": 3.7353,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22294696608867e-05,
+      "loss": 3.7346,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222108371337618e-05,
+      "loss": 3.7362,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221269776586566e-05,
+      "loss": 3.7235,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220431181835514e-05,
+      "loss": 3.7322,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219592587084462e-05,
+      "loss": 3.7252,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2187556302137833e-05,
+      "loss": 3.7171,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217917035462731e-05,
+      "loss": 3.7147,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217080078592052e-05,
+      "loss": 3.7215,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216241483841e-05,
+      "loss": 3.7196,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2154028890899476e-05,
+      "loss": 3.7357,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145642943388956e-05,
+      "loss": 3.7276,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2137256995878436e-05,
+      "loss": 3.7161,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128871048367915e-05,
+      "loss": 3.737,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2120485100857395e-05,
+      "loss": 3.7215,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211209915334688e-05,
+      "loss": 3.7336,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210372958464009e-05,
+      "loss": 3.7219,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209534363712957e-05,
+      "loss": 3.7073,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208695768961905e-05,
+      "loss": 3.7352,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207857174210853e-05,
+      "loss": 3.7185,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207020217340174e-05,
+      "loss": 3.7299,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206181622589122e-05,
+      "loss": 3.7192,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20534302783807e-05,
+      "loss": 3.7183,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204506070967391e-05,
+      "loss": 3.7077,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203667476216339e-05,
+      "loss": 3.7196,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202828881465287e-05,
+      "loss": 3.7304,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201990286714235e-05,
+      "loss": 3.7152,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011516919631836e-05,
+      "loss": 3.7294,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2003130972121316e-05,
+      "loss": 3.7196,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994745024610796e-05,
+      "loss": 3.7114,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1986359077100276e-05,
+      "loss": 3.7338,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1977989508393485e-05,
+      "loss": 3.7046,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1969603560882965e-05,
+      "loss": 3.7008,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1961233992176174e-05,
+      "loss": 3.7183,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1952848044665654e-05,
+      "loss": 3.7267,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1944462097155134e-05,
+      "loss": 3.7055,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1936076149644614e-05,
+      "loss": 3.7227,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927690202134094e-05,
+      "loss": 3.7028,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1919304254623574e-05,
+      "loss": 3.7155,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910918307113054e-05,
+      "loss": 3.7223,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190254873840627e-05,
+      "loss": 3.7176,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189417916969948e-05,
+      "loss": 3.7205,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188579322218896e-05,
+      "loss": 3.7166,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187740727467844e-05,
+      "loss": 3.7276,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186902132716792e-05,
+      "loss": 3.7083,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18606353796574e-05,
+      "loss": 3.7194,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185224943214688e-05,
+      "loss": 3.7143,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184386348463636e-05,
+      "loss": 3.7038,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183547753712584e-05,
+      "loss": 3.717,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182710796841905e-05,
+      "loss": 3.7149,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818738399712256e-05,
+      "loss": 3.724,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181035245220174e-05,
+      "loss": 3.722,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180196650469122e-05,
+      "loss": 3.7004,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17935805571807e-05,
+      "loss": 3.7016,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178519460967018e-05,
+      "loss": 3.7177,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177680866215966e-05,
+      "loss": 3.7241,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1768422714649136e-05,
+      "loss": 3.713,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1760036767138616e-05,
+      "loss": 3.7085,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175166719843183e-05,
+      "loss": 3.7145,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174328125092131e-05,
+      "loss": 3.7132,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173491168221452e-05,
+      "loss": 3.7116,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1726525734704e-05,
+      "loss": 3.7021,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171813978719348e-05,
+      "loss": 3.7202,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170975383968296e-05,
+      "loss": 3.719,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170136789217244e-05,
+      "loss": 3.7189,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169298194466192e-05,
+      "loss": 3.7035,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16845959971514e-05,
+      "loss": 3.7072,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167621004964088e-05,
+      "loss": 3.7062,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166784048093409e-05,
+      "loss": 3.7154,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165945453342357e-05,
+      "loss": 3.711,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1651084964716786e-05,
+      "loss": 3.7073,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164269901720626e-05,
+      "loss": 3.7103,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163431306969574e-05,
+      "loss": 3.7028,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162592712218522e-05,
+      "loss": 3.7063,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16175411746747e-05,
+      "loss": 3.701,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160915522716418e-05,
+      "loss": 3.7029,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160076927965366e-05,
+      "loss": 3.7067,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1592383332143145e-05,
+      "loss": 3.7142,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1584013763436354e-05,
+      "loss": 3.7001,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575627815925834e-05,
+      "loss": 3.7037,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1567241868415314e-05,
+      "loss": 3.7052,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558872299708523e-05,
+      "loss": 3.7122,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1550486352198003e-05,
+      "loss": 3.7031,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154210040468748e-05,
+      "loss": 3.7168,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153371445717696e-05,
+      "loss": 3.7125,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152532850966644e-05,
+      "loss": 3.7125,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151695894095965e-05,
+      "loss": 3.7083,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150857299344913e-05,
+      "loss": 3.7069,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150018704593861e-05,
+      "loss": 3.7059,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14918010984281e-05,
+      "loss": 3.708,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148341515091758e-05,
+      "loss": 3.7052,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147502920340706e-05,
+      "loss": 3.7004,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146664325589654e-05,
+      "loss": 3.711,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145827368718975e-05,
+      "loss": 3.7077,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144990411848296e-05,
+      "loss": 3.7015,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144151817097244e-05,
+      "loss": 3.6924,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143313222346192e-05,
+      "loss": 3.6987,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14247462759514e-05,
+      "loss": 3.6966,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141636032844088e-05,
+      "loss": 3.7076,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140797438093036e-05,
+      "loss": 3.7124,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139958843341984e-05,
+      "loss": 3.7194,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139120248590932e-05,
+      "loss": 3.6918,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138283291720253e-05,
+      "loss": 3.7113,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137444696969201e-05,
+      "loss": 3.7062,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136606102218149e-05,
+      "loss": 3.7071,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135767507467097e-05,
+      "loss": 3.6971,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1349289127160446e-05,
+      "loss": 3.7106,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340903179649926e-05,
+      "loss": 3.6979,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332517232139406e-05,
+      "loss": 3.7156,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132414766343262e-05,
+      "loss": 3.695,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315761715922095e-05,
+      "loss": 3.7082,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1307375768411575e-05,
+      "loss": 3.705,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298989820901055e-05,
+      "loss": 3.7038,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129062025219427e-05,
+      "loss": 3.7084,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128223430468375e-05,
+      "loss": 3.7072,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127384835717323e-05,
+      "loss": 3.7069,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126546240966271e-05,
+      "loss": 3.6972,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125707646215219e-05,
+      "loss": 3.7122,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8773159980773926,
+      "eval_runtime": 302.013,
+      "eval_samples_per_second": 1263.492,
+      "eval_steps_per_second": 39.485,
+      "step": 534240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124869051464167e-05,
+      "loss": 3.695,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124030456713115e-05,
+      "loss": 3.6872,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123191861962063e-05,
+      "loss": 3.705,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122353267211011e-05,
+      "loss": 3.7082,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121514672459959e-05,
+      "loss": 3.7095,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120676077708907e-05,
+      "loss": 3.6945,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119839120838228e-05,
+      "loss": 3.7007,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119000526087176e-05,
+      "loss": 3.6884,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118161931336124e-05,
+      "loss": 3.706,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117323336585072e-05,
+      "loss": 3.6989,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11648474183402e-05,
+      "loss": 3.7013,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1156461470829686e-05,
+      "loss": 3.7014,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1148091902122895e-05,
+      "loss": 3.6973,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1139705954612375e-05,
+      "loss": 3.695,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1131320007101855e-05,
+      "loss": 3.6902,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1122934059591335e-05,
+      "loss": 3.6841,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111454811208081e-05,
+      "loss": 3.6926,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110616216457029e-05,
+      "loss": 3.7014,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109777621705977e-05,
+      "loss": 3.695,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108939026954925e-05,
+      "loss": 3.7054,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108102070084246e-05,
+      "loss": 3.7104,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.107263475333194e-05,
+      "loss": 3.7082,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.106426518462515e-05,
+      "loss": 3.6992,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.105587923711463e-05,
+      "loss": 3.6972,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.104749328960411e-05,
+      "loss": 3.6953,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103910734209359e-05,
+      "loss": 3.7045,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.103072139458307e-05,
+      "loss": 3.695,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.102233544707255e-05,
+      "loss": 3.6939,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.101394949956203e-05,
+      "loss": 3.6928,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.100556355205151e-05,
+      "loss": 3.6889,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.099719398334472e-05,
+      "loss": 3.6888,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098882441463793e-05,
+      "loss": 3.7001,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.098043846712741e-05,
+      "loss": 3.7019,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097205251961689e-05,
+      "loss": 3.7015,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096366657210638e-05,
+      "loss": 3.687,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095528062459586e-05,
+      "loss": 3.7012,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.094689467708534e-05,
+      "loss": 3.695,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093850872957482e-05,
+      "loss": 3.6822,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0930139160868026e-05,
+      "loss": 3.6806,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0921769592161235e-05,
+      "loss": 3.6865,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0913383644650715e-05,
+      "loss": 3.6898,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0904997697140195e-05,
+      "loss": 3.6957,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0896611749629675e-05,
+      "loss": 3.6966,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0888225802119155e-05,
+      "loss": 3.6815,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0879839854608635e-05,
+      "loss": 3.703,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0871453907098115e-05,
+      "loss": 3.6892,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0863067959587595e-05,
+      "loss": 3.7011,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.085469839088081e-05,
+      "loss": 3.6921,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.084631244337029e-05,
+      "loss": 3.6686,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.083792649585977e-05,
+      "loss": 3.7055,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082954054834925e-05,
+      "loss": 3.6827,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.082117097964246e-05,
+      "loss": 3.6951,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081278503213194e-05,
+      "loss": 3.6858,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080439908462142e-05,
+      "loss": 3.6858,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.079602951591463e-05,
+      "loss": 3.6796,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078764356840411e-05,
+      "loss": 3.6819,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077925762089359e-05,
+      "loss": 3.7019,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077087167338307e-05,
+      "loss": 3.6762,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076248572587255e-05,
+      "loss": 3.6972,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075409977836203e-05,
+      "loss": 3.6895,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0745713830851515e-05,
+      "loss": 3.6757,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0737344262144724e-05,
+      "loss": 3.6975,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0728958314634204e-05,
+      "loss": 3.6784,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0720572367123684e-05,
+      "loss": 3.6706,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0712186419613164e-05,
+      "loss": 3.6825,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0703800472102644e-05,
+      "loss": 3.6909,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069541452459212e-05,
+      "loss": 3.674,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06870285770816e-05,
+      "loss": 3.6897,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067864262957108e-05,
+      "loss": 3.672,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0670273060864286e-05,
+      "loss": 3.6815,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0661887113353766e-05,
+      "loss": 3.6892,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065350116584325e-05,
+      "loss": 3.6846,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064511521833273e-05,
+      "loss": 3.6928,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.063674564962594e-05,
+      "loss": 3.6811,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.062835970211542e-05,
+      "loss": 3.6925,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06199737546049e-05,
+      "loss": 3.6798,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.061158780709438e-05,
+      "loss": 3.6874,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.060320185958386e-05,
+      "loss": 3.6764,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.059481591207334e-05,
+      "loss": 3.6759,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058642996456282e-05,
+      "loss": 3.6835,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05780440170523e-05,
+      "loss": 3.6832,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056967444834551e-05,
+      "loss": 3.6895,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056128850083499e-05,
+      "loss": 3.6926,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055290255332447e-05,
+      "loss": 3.6663,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054451660581395e-05,
+      "loss": 3.6712,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0536147037107167e-05,
+      "loss": 3.6819,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0527777468400376e-05,
+      "loss": 3.6919,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0519391520889856e-05,
+      "loss": 3.6817,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0511005573379336e-05,
+      "loss": 3.6759,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0502619625868816e-05,
+      "loss": 3.6872,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0494233678358296e-05,
+      "loss": 3.6814,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0485864109651505e-05,
+      "loss": 3.68,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0477478162140985e-05,
+      "loss": 3.6721,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0469092214630465e-05,
+      "loss": 3.6839,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0460706267119944e-05,
+      "loss": 3.6902,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0452320319609424e-05,
+      "loss": 3.6867,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0443934372098904e-05,
+      "loss": 3.6713,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0435548424588384e-05,
+      "loss": 3.673,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.042716247707787e-05,
+      "loss": 3.6729,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041879290837108e-05,
+      "loss": 3.6852,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.041040696086056e-05,
+      "loss": 3.6804,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.040203739215377e-05,
+      "loss": 3.6759,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039365144464325e-05,
+      "loss": 3.678,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038526549713273e-05,
+      "loss": 3.6725,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037687954962221e-05,
+      "loss": 3.674,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036849360211169e-05,
+      "loss": 3.6675,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036010765460117e-05,
+      "loss": 3.6738,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035172170709065e-05,
+      "loss": 3.6757,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034333575958012e-05,
+      "loss": 3.6812,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033496619087334e-05,
+      "loss": 3.671,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0326580243362825e-05,
+      "loss": 3.6651,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0318194295852305e-05,
+      "loss": 3.6773,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0309824727145514e-05,
+      "loss": 3.6791,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0301438779634994e-05,
+      "loss": 3.6724,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0293052832124474e-05,
+      "loss": 3.6817,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0284666884613954e-05,
+      "loss": 3.6802,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027628093710343e-05,
+      "loss": 3.687,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026791136839664e-05,
+      "loss": 3.6742,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025952542088612e-05,
+      "loss": 3.6758,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0251139473375596e-05,
+      "loss": 3.6716,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0242753525865076e-05,
+      "loss": 3.683,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.023436757835456e-05,
+      "loss": 3.6747,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.022598163084404e-05,
+      "loss": 3.6663,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021759568333352e-05,
+      "loss": 3.678,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020922611462673e-05,
+      "loss": 3.6824,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020084016711621e-05,
+      "loss": 3.6672,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.019247059840943e-05,
+      "loss": 3.6651,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01840846508989e-05,
+      "loss": 3.6685,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.017569870338838e-05,
+      "loss": 3.6673,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.016731275587786e-05,
+      "loss": 3.6756,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015892680836734e-05,
+      "loss": 3.6838,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.015055723966055e-05,
+      "loss": 3.6867,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014217129215003e-05,
+      "loss": 3.6608,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0133785344639516e-05,
+      "loss": 3.6826,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0125399397128996e-05,
+      "loss": 3.6748,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0117013449618476e-05,
+      "loss": 3.6739,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0108627502107956e-05,
+      "loss": 3.666,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0100241554597436e-05,
+      "loss": 3.6768,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0091855607086916e-05,
+      "loss": 3.6666,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0083486038380125e-05,
+      "loss": 3.6848,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0075100090869605e-05,
+      "loss": 3.6663,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0066714143359085e-05,
+      "loss": 3.6766,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0058328195848565e-05,
+      "loss": 3.6734,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0049958627141774e-05,
+      "loss": 3.677,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.004158905843498e-05,
+      "loss": 3.6751,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.003320311092447e-05,
+      "loss": 3.6753,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.002481716341395e-05,
+      "loss": 3.6787,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.001643121590343e-05,
+      "loss": 3.6676,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000804526839291e-05,
+      "loss": 3.6801,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8724730014801025,
+      "eval_runtime": 303.6358,
+      "eval_samples_per_second": 1256.739,
+      "eval_steps_per_second": 39.274,
+      "step": 610560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999965932088239e-05,
+      "loss": 3.6657,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999127337337187e-05,
+      "loss": 3.6596,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.998288742586135e-05,
+      "loss": 3.6727,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997451785715456e-05,
+      "loss": 3.6769,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.996613190964404e-05,
+      "loss": 3.6815,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.995774596213352e-05,
+      "loss": 3.6646,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994937639342673e-05,
+      "loss": 3.6752,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994099044591621e-05,
+      "loss": 3.6612,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.993260449840569e-05,
+      "loss": 3.6707,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.992421855089517e-05,
+      "loss": 3.6694,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.991583260338465e-05,
+      "loss": 3.6718,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9907446655874134e-05,
+      "loss": 3.6694,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989907708716734e-05,
+      "loss": 3.671,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989069113965682e-05,
+      "loss": 3.6632,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98823051921463e-05,
+      "loss": 3.6622,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987391924463578e-05,
+      "loss": 3.6537,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9865533297125256e-05,
+      "loss": 3.6637,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9857147349614736e-05,
+      "loss": 3.6737,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984877778090795e-05,
+      "loss": 3.6622,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984039183339743e-05,
+      "loss": 3.676,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9832005885886905e-05,
+      "loss": 3.6796,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9823619938376385e-05,
+      "loss": 3.6761,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.981523399086587e-05,
+      "loss": 3.6746,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.980686442215909e-05,
+      "loss": 3.6651,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979847847464856e-05,
+      "loss": 3.6644,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.979009252713804e-05,
+      "loss": 3.6727,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.978170657962752e-05,
+      "loss": 3.6715,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9773320632117e-05,
+      "loss": 3.665,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.976493468460648e-05,
+      "loss": 3.6635,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.975656511589969e-05,
+      "loss": 3.6598,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974817916838917e-05,
+      "loss": 3.6563,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973979322087865e-05,
+      "loss": 3.6719,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.973140727336813e-05,
+      "loss": 3.6736,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.972302132585761e-05,
+      "loss": 3.6722,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.971463537834709e-05,
+      "loss": 3.6597,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.970624943083657e-05,
+      "loss": 3.6689,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9697879862129786e-05,
+      "loss": 3.6674,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9689493914619266e-05,
+      "loss": 3.6517,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9681107967108746e-05,
+      "loss": 3.6561,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9672722019598225e-05,
+      "loss": 3.6551,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9664368829695164e-05,
+      "loss": 3.6596,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9655982882184644e-05,
+      "loss": 3.6709,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9647596934674124e-05,
+      "loss": 3.6653,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9639210987163604e-05,
+      "loss": 3.6545,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9630825039653084e-05,
+      "loss": 3.6705,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9622439092142564e-05,
+      "loss": 3.659,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9614053144632043e-05,
+      "loss": 3.6734,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.960568357592526e-05,
+      "loss": 3.6635,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.959729762841474e-05,
+      "loss": 3.6389,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958891168090422e-05,
+      "loss": 3.6764,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.95805257333937e-05,
+      "loss": 3.656,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957213978588318e-05,
+      "loss": 3.6652,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956375383837266e-05,
+      "loss": 3.6588,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.955536789086214e-05,
+      "loss": 3.6551,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.954699832215535e-05,
+      "loss": 3.6507,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953861237464483e-05,
+      "loss": 3.6562,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953022642713431e-05,
+      "loss": 3.6739,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952184047962379e-05,
+      "loss": 3.6466,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.951345453211327e-05,
+      "loss": 3.6713,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950506858460275e-05,
+      "loss": 3.6599,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.949668263709223e-05,
+      "loss": 3.6439,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9488313068385444e-05,
+      "loss": 3.6697,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947992712087492e-05,
+      "loss": 3.6512,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94715411733644e-05,
+      "loss": 3.6395,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.946315522585388e-05,
+      "loss": 3.6543,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945476927834336e-05,
+      "loss": 3.6624,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.944638333083284e-05,
+      "loss": 3.6482,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.943799738332232e-05,
+      "loss": 3.6599,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94296114358118e-05,
+      "loss": 3.646,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9421241867105006e-05,
+      "loss": 3.6498,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9412855919594486e-05,
+      "loss": 3.6615,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94044863508877e-05,
+      "loss": 3.6581,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.939610040337718e-05,
+      "loss": 3.6605,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.938771445586666e-05,
+      "loss": 3.6556,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937932850835614e-05,
+      "loss": 3.6682,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.937094256084562e-05,
+      "loss": 3.6492,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.93625566133351e-05,
+      "loss": 3.6592,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.935418704462831e-05,
+      "loss": 3.6472,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.934580109711779e-05,
+      "loss": 3.6485,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.933741514960727e-05,
+      "loss": 3.6579,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932902920209675e-05,
+      "loss": 3.6518,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932064325458623e-05,
+      "loss": 3.6615,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.931225730707571e-05,
+      "loss": 3.667,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.930388773836892e-05,
+      "loss": 3.6384,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.92955017908584e-05,
+      "loss": 3.642,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928711584334788e-05,
+      "loss": 3.6562,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9278729895837366e-05,
+      "loss": 3.6589,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9270343948326846e-05,
+      "loss": 3.657,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9261958000816326e-05,
+      "loss": 3.6492,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9253588432109535e-05,
+      "loss": 3.657,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9245202484599015e-05,
+      "loss": 3.6528,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9236816537088495e-05,
+      "loss": 3.6489,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9228430589577975e-05,
+      "loss": 3.6487,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9220044642067455e-05,
+      "loss": 3.6575,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9211675073360664e-05,
+      "loss": 3.6641,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9203289125850144e-05,
+      "loss": 3.6579,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9194903178339624e-05,
+      "loss": 3.6421,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9186517230829104e-05,
+      "loss": 3.6476,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917814766212232e-05,
+      "loss": 3.6462,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91697617146118e-05,
+      "loss": 3.6554,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.916137576710128e-05,
+      "loss": 3.6537,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.915298981959075e-05,
+      "loss": 3.6467,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.914460387208023e-05,
+      "loss": 3.6508,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.913621792456971e-05,
+      "loss": 3.6428,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.912783197705919e-05,
+      "loss": 3.6473,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911944602954867e-05,
+      "loss": 3.6411,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911107646084188e-05,
+      "loss": 3.6486,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.910269051333136e-05,
+      "loss": 3.65,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.909430456582084e-05,
+      "loss": 3.653,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.908591861831032e-05,
+      "loss": 3.6419,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90775326707998e-05,
+      "loss": 3.6441,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906914672328929e-05,
+      "loss": 3.6465,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90607771545825e-05,
+      "loss": 3.657,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.905239120707198e-05,
+      "loss": 3.6461,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904400525956146e-05,
+      "loss": 3.6495,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.903561931205094e-05,
+      "loss": 3.652,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.902723336454042e-05,
+      "loss": 3.6645,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90188474170299e-05,
+      "loss": 3.6473,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.901046146951938e-05,
+      "loss": 3.6489,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9002091900812586e-05,
+      "loss": 3.6483,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8993705953302066e-05,
+      "loss": 3.653,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8985320005791546e-05,
+      "loss": 3.6476,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8976934058281026e-05,
+      "loss": 3.641,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8968548110770506e-05,
+      "loss": 3.65,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8960162163259986e-05,
+      "loss": 3.6541,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8951776215749466e-05,
+      "loss": 3.6431,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.894340664704268e-05,
+      "loss": 3.6391,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.893502069953216e-05,
+      "loss": 3.6416,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.892663475202164e-05,
+      "loss": 3.6376,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8918248804511115e-05,
+      "loss": 3.6522,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.890987923580433e-05,
+      "loss": 3.6528,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.890149328829381e-05,
+      "loss": 3.6601,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.889310734078329e-05,
+      "loss": 3.6357,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8884721393272764e-05,
+      "loss": 3.6566,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887635182456598e-05,
+      "loss": 3.6496,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886796587705546e-05,
+      "loss": 3.6486,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885957992954494e-05,
+      "loss": 3.6423,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885119398203442e-05,
+      "loss": 3.6439,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.88428080345239e-05,
+      "loss": 3.6419,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.883442208701338e-05,
+      "loss": 3.6585,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882603613950286e-05,
+      "loss": 3.6369,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.881765019199234e-05,
+      "loss": 3.6514,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880928062328555e-05,
+      "loss": 3.6503,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880089467577503e-05,
+      "loss": 3.6529,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.879252510706824e-05,
+      "loss": 3.6408,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878413915955772e-05,
+      "loss": 3.6497,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.87757532120472e-05,
+      "loss": 3.6518,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876736726453668e-05,
+      "loss": 3.6417,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.875899769582989e-05,
+      "loss": 3.6523,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.868178129196167,
+      "eval_runtime": 303.4461,
+      "eval_samples_per_second": 1257.525,
+      "eval_steps_per_second": 39.299,
+      "step": 686880
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 4.63322220824107e+17,
+  "trial_name": null,
+  "trial_params": null
+}